1. Field of the Invention
The present invention relates to a flow meter for measuring the flow of very high void fraction multi-phase fluid streams. More particularly, the present invention relates to an apparatus and method in which multiple pressure differentials are used to determine flow rates of gas and liquid phases of a predominantly gas fluid stream to thereby determine the mass flow rate of each phase.
2. State of the Art
There are many situations where it is desirable to monitor multi-phase fluid streams prior to separation. For example, in oil well or gas well management, it is important to know the relative quantities of gas and liquid in a multi-phase fluid stream, to thereby enable determination of the amount of gas, etc. actually obtained. This is of critical importance in situations, such as off-shore drilling, in which it is common for the production lines of several different companies to be tied into a common distribution line to carry the fuel back to shore. While a common method for metering a gas is to separate out the liquid phase, such a system in not desirable for fiscal reasons. When multiple production lines feed into a common distribution line, it is important to know the flow rates from each production line to thereby provide an accurate accounting for the production facilities.
In recent years, the metering of multi-phase fluid streams prior to separation has achieved increased attention. Significant progress has been made in the metering of multi-phase fluids by first homogenizing the flow in a mixer then metering the pseudo single phase fluid in a venturi in concert with a gamma densitometer or similar device. This approach relies on the successful creation of a homogenous mixture with equal phase velocities, which behaves as if it were a single phase fluid with mixture density {overscore (xcfx81)}=xcex1xcfx81g+(1xe2x88x92xcex1)xcfx81l where xcex1 is the volume fraction of the gas phase, and xcfx81g is the gas phase density and xcfx811 is the liquid phase density. This technique works well for flows which after homogenizing the continuous phase is a liquid phase. While the upper limit of applicability of this approach is ill defined, it is generally agreed that for void fractions greater than about ninety to ninety-five percent (90-95%) a homogenous mixture is difficult to create or sustain. The characteristic undisturbed flow regime in this void fraction range is that of an annular or ring shaped flow configuration. The gas phase flows in the center of the channel and the liquid phase adheres to and travels along the sidewall of the conduit as a thick film. Depending on the relative flow rates of each phase, significant amounts of the denser phase may also become entrained in the gas phase and be conveyed as dispersed droplets. Nonetheless, a liquid film is always present on the wall of the conduit. While the liquid generally occupies less than five percent (5%) of the cross-sectional volume of the flow channel, the mass flow rate of the liquid may be comparable to or even several times greater than that of the gas phase due to its greater density.
The fact that the phases are partially or fully separated, and consequently have phase velocities which are significantly different (slip), complicates the metering problem. The presence of the liquid phase distorts the gas mass flow rate measurements and causes conventional meters, such as orifice plates or venturi meters, to overestimate the flow rate of the gas phase. For example the gas mass flow can be estimated using the standard equation                               m          g                =                                            A              ⁢                              xe2x80x83                            ⁢                              C                c                            ⁢              Y                                                      1                -                                  β                  4                                                              ⁢                                    2              ⁢                              ρ                g                            ⁢              Δ              ⁢                              xe2x80x83                            ⁢              P                                                          [        7        ]            
where mg is the gas mass flow rate, A is the area of the throat, xcex94P is the measured pressure differential, xcfx81g the gas density at flow conditions, Cc the discharge coefficient, and Y is the expansion factor. In test samples using void fractions ranging from 0.997 to 0.95, the error in the measured gas mass flow rate ranges from 7% to 30%. It is important to note that the presence of the liquid phase increases the pressure drop in the venturi and results in over-predicting the true gas mass flow rate. The pressure drop is caused by the interaction between the gas and liquid phases. Liquid droplet acceleration by the gas, irreversible drag force work done by the gas phase in accelerating the liquid film and wall losses determine the magnitude of the observed pressure drop. In addition, the flow is complicated by the continuous entrainment of liquid into the gas, the redeposition of liquid from the gas into the liquid film along the venturi length, and also by the presence of surface waves on the surface of the annular or ringed liquid phase film. The surface waves on the liquid create a roughened surface over which the gas must flow increasing the momentum loss due to the addition of drag at the liquid/gas interface.
Other simple solutions have been proposed to solve the overestimation of gas mass flow rate under multi-phase conditions. For example, Murdock, ignores any interaction (momentum exchange) between the gas and liquid phases and proposed to calculate the gas mass flow if the ratio of gas to liquid mass flow is known in advance. See Murdock, J. W. (1962). Two Phase Flow Measurement with Orifices, ASME Journal of Basic Engineering, December, 419-433. Unfortunately this method still has up to a 20% error rate or more.
While past attempts at metering multi-phase fluid streams have produced acceptable results below the ninety to ninety five percent (90-95%) void fraction range, they have not provided satisfactory metering for the very high void multi-phase flows which have less than five to ten (5-10%) non-gas phase by volume. When discussing large amounts of natural gas or other fuel, even a few percent difference in the amount of non-gas phase can mean substantial differences in the value of a production facility. For example, if there are two wells which produce equal amounts of natural gas per day. The first well produces, by volume, 1% liquid and the second well produces 5% liquid. If a conventional mass flow rate meter is relied upon to determine the amount of gas produced, the second well will erroneously appear to produce as much as 20-30% more gas than the first well. Suppose further that the liquid produced is a light hydrocarbon liquid (e.g. a gas condensate such as butane or propane) which is valuable in addition to the natural gas produced. Conventional meters will provide no information about the amount of liquid produced. Then if the amount of liquid produced is equally divided between the two wells, the value of the production from the first well will be overestimated while the production from the second well will be underestimated. To properly value the gas and liquid production from both wells, a method of more accurately determining the mass flow rate of both the gas and liquid phases is required.
The prior art, however, has been generally incapable of accurately metering the very high void multi-phase fluid streams. In light of the problems of the prior art, there is a need for an apparatus and method that is less complex and provides increased accuracy for very high void multi-phase fluid streams. Such an apparatus and method should be physically rugged, simple to use, and less expensive than current technology.
It is an object of the present invention to provide an improved apparatus and method for metering very high void multi-phase fluid streams.
It is another object of the present invention to provide an apparatus and method which increases the accuracy of metering with respect to both the gas phase and the liquid phase when measuring very high void multi-phase fluid streams.
It is still another object of the present invention to provide such an apparatus and method which does not require homogenization or separation of the multi-phase fluid in order to determine flow rate for each of the phases.
The above and other objects of the invention are realized in a specific method for metering the phases of a multiple phase fluid. The flow meter includes a cross-sectional area change in the flow conduit such as a venturi with an elongate passage. Disposed along the elongate passage is a converging section, an extended throat section, and a diffuser. The flow meter also includes a plurality of pressure monitoring sites which are used to monitor pressure changes which occur as the multi-phase fluid passes through the elongate passage. These pressure changes, in turn, can be processed to provide information as to the respective flow rates of the phases of the multi-phase fluid. By determining the flow rates of the components of the multi-phase fluid, the amount of natural gas, etc., can be accurately determined and accounting improved.
In accordance with another aspect of the present invention a method for determining the mass flow of the high void fraction fluid flow and the gas flow includes a number of steps. The first step is calculating a gas density for the gas flow. The next two steps are finding the normalized gas mass flow rate through the venturi and then computing the actual gas mass flow rate. The following step is estimating the gas velocity in the venturi tube throat. The next step is calculating the additional pressure drop experienced by the gas phase due to work performed by the gas phase in accelerating the liquid phase between the upstream pressure measuring point and the pressure measuring point at the end of the venturi contraction or throat. Yet another step is estimating the liquid velocity in the venturi throat using the calculated pressure drop experienced by the gas-phase due to work performed by the gas phase. Then, the friction loss is computed between the liquid phase and the conduit wall in the venturi tube using the liquid velocity. Finally, the total mass flow rate based on measured pressure in the venturi throat is calculated, and the liquid mass flow rate is calculated by subtracting the total mass flow rate and the gas mass flow rate.